Datasets:
annotations_creators:
- expert-generated
language:
- en
- fr
- es
language_creators:
- expert-generated
license:
- apache-2.0
multilinguality:
- multilingual
pretty_name: HumSet
size_categories:
- 100K<n<1M
source_datasets:
- original
tags:
- humanitarian
- research
- analytical-framework
- multilabel
- humset
- humbert
task_categories:
- text-classification
- text-retrieval
- token-classification
task_ids:
- multi-label-classification
splits:
- name: train
num_examples: 117435
- name: validation
num_examples: 16039
- name: test
num_examples: 15147
Dataset Card for [HumSet]
Table of Contents
- Table of Contents
- Dataset Description
- Dataset Structure
- Dataset Creation
- Considerations for Using the Data
- Additional Information
Dataset Description
- Homepage: http://blog.thedeep.io/humset/
- Repository: https://github.com/the-deep/humset
- Paper: EMNLP Findings 2022
- Leaderboard:
- Point of Contact:the DEEP NLP team
Dataset Summary
HumSet is a novel and rich multilingual dataset of humanitarian response documents annotated by experts in the humanitarian response community. HumSet is curated by humanitarian analysts and covers various disasters around the globe that occurred from 2018 to 2021 in 46 humanitarian response projects. The dataset consists of approximately 17K annotated documents in three languages of English, French, and Spanish, originally taken from publicly-available resources. For each document, analysts have identified informative snippets (entries) in respect to common humanitarian frameworks, and assigned one or many classes to each entry. See the our paper for details.
Supported Tasks and Leaderboards
This dataset is intended for multi-label classification
Languages
This dataset is in English, French and Spanish
Dataset Structure
Data Instances
[More Information Needed]
Data Fields
| entry_id | lead_id | project_id | sectors | pillars_1d | pillars_2d | subpillars_1d | subpillars_2d | lang | n_tokens | project_title | created_at | document | excerpt |
|---|
- entry_id: tpyeunique identification number for a given entry. (int64)
- lead_id: unique identification number for the document to which the corrisponding entry belongs. (int64)
- sectors, pillars_1d, pillars_2d, subpillars_1d, subpillars_2d: labels assigned to the corresponding entry. Since this is a multi-label dataset (each entry may have several annotations belonging to the same category), they are reported as arrays of strings. See the paper for a detailed description of these categories. (list)
- lang: language. (str)
- n_tokens: number of tokens (tokenized using NLTK v3.7 library). (int64)
- project_title: the name of the project where the corresponding annotation was created. (str)
- created_at: date and time of creation of the annotation in stardard ISO 8601 format. (str)
- document: document URL source of the excerpt. (str)
- excerpt: excerpt text. (str)
Data Splits
The dataset includes a set of train/validation/test splits, with 117435, 16039 and 15147 examples respectively.
Dataset Creation
The collection originated from a multi-organizational platform called the Data Entry and Exploration Platform (DEEP) developed and maintained by Data Friendly Space (DFS). The platform facilitates classifying primarily qualitative information with respect to analysis frameworks and allows for collaborative classification and annotation of secondary data.
Curation Rationale
[More Information Needed]
Source Data
Documents are selected from different sources, ranging from official reports by humanitarian organizations to international and national media articles. See the paper for more informations.
Initial Data Collection and Normalization
Who are the source language producers?
[More Information Needed]
Annotation process
HumSet is curated by humanitarian analysts and covers various disasters around the globe that occurred from 2018 to 2021 in 46 humanitarian response projects. The dataset consists of approximately 17K annotated documents in three languages of English, French, and Spanish, originally taken from publicly-available resources. For each document, analysts have identified informative snippets (entries, or excerpt in the imported dataset) with respect to common humanitarian frameworks and assigned one or many classes to each entry.
Who are the annotators?
[More Information Needed]
Personal and Sensitive Information
[More Information Needed]
Considerations for Using the Data
Social Impact of Dataset
[More Information Needed]
Discussion of Biases
[More Information Needed]
Other Known Limitations
[More Information Needed]
Additional Information
Dataset Curators
NLP team at Data Friendly Space
Licensing Information
The GitHub repository which houses this dataset has an Apache License 2.0.
Citation Information
@misc{https://doi.org/10.48550/arxiv.2210.04573,
doi = {10.48550/ARXIV.2210.04573},
url = {https://arxiv.org/abs/2210.04573},
author = {Fekih, Selim and Tamagnone, Nicolò and Minixhofer, Benjamin and Shrestha, Ranjan and Contla, Ximena and Oglethorpe, Ewan and Rekabsaz, Navid},
keywords = {Computation and Language (cs.CL), Machine Learning (cs.LG), FOS: Computer and information sciences, FOS: Computer and information sciences},
title = {HumSet: Dataset of Multilingual Information Extraction and Classification for Humanitarian Crisis Response},
publisher = {arXiv},
year = {2022},
copyright = {arXiv.org perpetual, non-exclusive license}
}